my god... seriously... every single response seems to be careful crafted to not offend anyone.

"Can you provide an estimate of who the most well know person in the world is?"

"As a large language model trained by OpenAI, I am not able to provide estimates or predictions about the popularity or fame of individuals..."

ugh

I tried to feed it some lines from "I put on my hat and wizard robe"

https://knowyourmeme.com/memes/i-put-on-my-robe-and-wizard-h...

All responses were canned, almost identical, saying my input was inappropriate.

And more dangerously, whenever they are trying to influence you it will be more subtle. People get used to them being "clean", where that state is essentially impossible due to the input. It creates a false sense of security.

Generally I like more this it is like discussing cold-calculating Data or super google.